According to news on June 5, Microsoft’s Scout AI assistant had just been released, but it caused controversy due to an internal strategy document: the document stated the first-phase goal as “making people addicted” (Make people addicted). Later, Microsoft CEO Satya Nadella expressed opposition to this route internally.
According to an internal document obtained by 404 Media, the material is called "ClawPilot: Overview and "Project Lobster" Planning" and designs a three-stage advancement path for Scout. According to the document, Scout is an "always-on" personal AI agent integrated into Microsoft 365 that can handle calendar, email, meeting scheduling and other work tasks.
The controversy comes from "people can't live without it every day"
The most glaring thing in the document is the goal of the first phase: "to make people addicted." The document further explains that at this stage, we will continue to deliver an independent ClawPilot experience, polish the user experience, expand the user base, and build an ecosystem of skills and tools so that people can’t live without it every day.
404 Media reported that AI was also involved in writing the document itself. The document also said that Scout has become "one of the most popular tools" within Microsoft, used by more than 1,000 employees, including Nadella himself.
An anonymous Microsoft employee told 404 Media that writing "addiction" into the product strategy is "very disturbing." His concern is not simply a matter of wording, but that AI chatbots and AI agents are increasingly entering users' work, emotions, and daily decision-making scenarios. If a product actively pursues a sense of dependence, its boundaries will be more sensitive than ordinary software.
Is Nadella objecting to product direction or internal wording?
The Information later reported that Nadella pushed back on this line put forward by Microsoft executives. Publicly available information shows that Nadella’s opposition focuses on product plans that “make users addicted”, but the outside world is currently unable to confirm his original words internally, the scope of his delivery, and whether he named the specific person in charge.
In Microsoft’s official blog announcing Scout on June 2, there was no “addiction” related expression. The official statement emphasizes that Scout "understands how you work," takes action on things that users care about, and is "under your control." The blog also mentioned that Scout has built-in enterprise-level security and compliance controls, including Microsoft Entra identity management, Microsoft Purview data protection policies, and the need for manual approval for sensitive operations.
However, many foreign media quoted the 404 Media report as saying that the leaked document lists "security and compliance" as issues that need to be gradually resolved in the future, rather than as a prerequisite before release. For an office AI assistant that needs to access sensitive information such as email, calendar, and contacts, this sequence has a more direct impact on user trust than the "addictive" wording itself.
What Office users really have to worry about
The reason Scout controversy spreads easily is because it is not an entertainment product, but an AI assistant embedded in the office entrance. Short videos, mobile games, and social media use mechanisms to increase user retention time, and many people have already expected it; but if a tool that helps users handle emails, calendars, and meetings takes "cannot live without it" as the first-stage goal, users will not only face an efficiency tool, but a re-bound workflow.
For Microsoft 365 users, the risk is not that Scout will definitely become "addictive", but that the dependency relationship may be written in advance by the product design: it first learns your work habits, then takes over more daily tasks, and then connects with more AI tools. Each step can increase the cost of not using it.
Nadella's rebuttal at least shows that Microsoft's top management has realized that the word "addiction" cannot be the public direction of AI office products. However, this document was circulated internally and was obtained by the media in the form of an executive memo, which also shows that Microsoft still needs to clarify the growth goals, security order and user boundaries of AI agents internally.